In recent years, the emergence of economic policy uncertainty has had a certain impact on stock returns. It can be used as a research object to explore whether economic policy uncertainty can significantly predict future stock returns. This paper studies the predictive ability of economic policy uncertainty based on relevant data from major industries in the United States from January 1985 to December 2023, including Food industry (Food), Retail Stores industry (Rtail), Fabricated Products industry (FabPr), Drugs, Soap, Perfumes, Tobacco industry (Cnsum), Construction and Construction Materials industry (Cnstr), Consumer Durables industry (Durbl), Automobiles industry (Cars), Textiles, Apparel & Footwear industry (Clths), Machinery and Business Equipment industry (Machn), Chemicals industry (Chems), and Transportation industry(Trans), Utilities industry (Utils), Steel Works Etc industry (Steel), Mining and Minerals industry(Mines), Oil and Petroleum Products industry (Oil), and Banks, Insurance Companies, and Other Financials industry (Finan). Through both in-sample and out-of-sample examines, it is found that economic policy uncertainty is a reliable indicator for predicting stock returns, and its predictive performance is better in Food industry, Retail Stores industry, Fabricated Products industry, Transportation industry, Construction and Construction Materials industry, Consumer Durables industry, Automobiles industry, Textiles, Apparel & Footwear industry, Machinery and Business Equipment industry, Chemicals industry, and Drugs, Soap, Perfumes, Tobacco industry than in Utilities industry, Steel Works Etc industry, Mining and Minerals industry, Oil and Petroleum Products industry, and Banks, Insurance Companies, and Other Financials industry. Meanwhile, by replacing the measurement indicator of economic policy uncertainty, conducting long-horizon forecasting, and using different evaluation periods for robustness checks, it is found that the research conclusions are robust.
Read full abstract